On Wed, Feb 3, 2010 at 1:49 PM, Alex Hunsaker <badalex@gmail.com> wrote:
> On Wed, Feb 3, 2010 at 11:43, Alex Hunsaker <badalex@gmail.com> wrote:
>> On Wed, Feb 3, 2010 at 11:28, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>>> Tim Bunce <Tim.Bunce@pobox.com> writes:
>>>> I do see a need for a GRANT check and I'm adding one now (based on
>>>> the code in CreateFunction() in functioncmds.c - thanks to RhodiumToad
>>>> on IRC for the pointer).
>>>
>>> What exactly are you proposing to check, and where, and what do you
>>> think that will fix?
>>
>> Non plperl ...
>
> Put another way:
> HEAD: only people with plperl granted can make functions to manipulate $_SHARED
> PATCH: anyone can set plperl.plperl_safe_init (but note not
> plperlu_init as its SUSER) and manipulate $_SHARED
> Proposed fix: only people with plperl granted can set
> plperl.plplerl_safe_init and hence restore HEAD behavior
I think we should just not have any unprivileged-user-settable GUCs
and then this problem goes away. Trying to test for GRANT privilege
on the appropriate language seems like a big kludge, and I am doubtful
that it's worth it.
...Robert